<!--  -->
# whole-brain_DeepDenoising
Deep denoising pushes the limit of functional data acquisition by recovering high SNR calcium traces from low SNR videos acquired using low laser power or smaller exposure time. Thus deep denoising enables faster and longer volumetric recordings. For more details, our [paper](https://www.nature.com/articles/s41467-022-32886-w) is now published.
<p align = "center"><b>Denoise whole-brain videos</b></p>
<!-- <img src = "https://github.com/shiveshc/whole-brain_DeepDenoising/blob/main/extra/wb_denoising.gif" width=100% align="center"> -->
<p align = "center"><b>Denoise mechanosensory neurites</b></p>
<!-- <img src = "https://github.com/shiveshc/whole-brain_DeepDenoising/blob/main/extra/neurite_denoising.gif" width=100% align="center"> -->
<p align = "center"><b>Recover high SNR calcium traces</b></p>
<!-- <img src = "https://github.com/shiveshc/whole-brain_DeepDenoising/blob/main/extra/denoise_traces.png" width=100% align="center"> -->
# Contents
1. [Installation](#installation)
2. [Additional system requirements](#additional-system-requirements)
3. [Train on new dataset](#train-on-new-dataset)
4. [Denoise calcium activity recordings](#denoise-calcium-activity-recordings)
5. [Denoise independent images](#denoise-independent-images)
6. [Run sample datasets](#run-sample-datasets)
# Installation
Installation steps tested for Windows 10 64-bit and Python 3.5
### 1. Download whole-brain_DeepDenoising repository
Open Git Bash terminal, navigate to desired location and clone repository using `git clone https://github.com/shiveshc/whole-brain_DeepDenoising.git`.
Or click on `Code` button on top right corner and `Download ZIP`.
### 2. Setting up venv and installing libraries
Open command line terminal as administrator and navigate to cloned repository path using `cd .\whole-brain_DeepDenoising`.
Next run following commands -
| Windows 10 | MacOS |
|------------|-------|
| 1. `python -m venv env` | 1. `python -m venv env` |
| 2. `env\Scripts\activate.bat` | 2. `source env/bin/activate` |
| 3. `python -m pip install --upgrade "pip < 21.0"` | 3. `python -m pip install --upgrade "pip < 21.0"` |
| 4. `pip install -r requirements.txt` | 4. `pip install -r requirements.txt` |
Installation should take 10-15 minutes.
Common installation errors -
1. pip version is not upgraded\
__solution :__ upgrade pip using `python -m pip install --upgrade pip`
2. pip version is not compatible with python version\
__solution :__ install suitable pip version e.g. pip version < 21.0 is compatible with Python 3.5.
# Additional system requirements
Model training using GPUs is much faster and thus preferred. To be able to use GPUs, suitable NVIDIA drivers and CUDA libraries must be installed. Additional instructions on setting up tensorflow-gpu can be found at <a href = "https://www.tensorflow.org/install/gpu#Software_requirements">Software requirements</a> and <a href = "https://www.tensorflow.org/install/gpu#windows_setup">Windows setup</a>.
For tensorflow 1.6.0 currently setup in venv, CUDA v9.0 and `cudnn-9.0-windows10-x64-v7.6.4.38` downloaded from <a href= "https://developer.nvidia.com/rdp/cudnn-archive">cudnn archives</a> works for Windows 10 64-bit.
# Train on new dataset
1. To train network on new data, pairs of noisy i.e. low SNR images (acquired at low laser power or small exposure time conditions) and clean i.e. high SNR images (acquired at high laser power or long exposure time conditions) are needed. Currently supported image size is 512 x 512 x d where d is number of images in stack. For other image sizes, either resize images first or channel dimensions need to be changed in architecture files in `cnn_archs` folder.
2. Structure of training data folder should be organised as below -
```
data
├───gt_imgs
│ ├───img_1
│ │ z_1.tif
│ │ z_2.tif
│ │ ...
│ │
│ ├───img_2
│ │ z_1.tif
│ │ z_2.tif
│ │ ...
│ │
│ └───img_3
│ z_1.tif
│ z_2.tif
│ ...
│
└───noisy_imgs
├───img_1
│ z_1.tif
│ z_2.tif
│ ...
│
├───img_2
│ z_1.tif
│ z_2.tif
│ ...
│
└───img_3
z_1.tif
z_2.tif
...
```
3. Run `python train.py -h` to see usage and input arguments. The output on terminal should look like below -
```
usage: train.py [-h] [-out OUT]
[-arch {unet,unet_fixed,hourglass_wres,hourglass_wores}]
[-mode {2D,2.5D,3D}] [-depth DEPTH] [-loss {l2,l1}]
[-epochs EPOCHS] [-lr LR] [-bs BS] [-tsize TSIZE]
data run {1,0}
train CNN model to denoise volumetric functional recordings
positional arguments:
data training data path
run run number to distinguish different runs
{1,0} 1 if train network on max projection of 3D stacks else
0
optional arguments:
-h, --help show this help message and exit
-out OUT location for saving results
-arch {unet,unet_fixed,hourglass_wres,hourglass_wores}
CNN architechture to use for training (default is
hourglass_wres)
-mode {2D,2.5D,3D} training mode (default is 2D)
-depth DEPTH stack depth to use for training (must be odd number,
default is 1)
-loss {l2,l1} L2 or L1 loss for training (default is l1)
-epochs EPOCHS number of epochs to train the model for (150-200 is
good choice, default is 150)
-lr LR learning rate (default is 0.001)
-bs BS batch size of training (default is 6)
-tsize TSIZE data size (number of images) to use for training
```
e.g. to train the network on whole-brain data with following settings -
- data path - `/training_data` (path should have forward slashes '/')
- run number - `1`
- train on max-projection (e.g. ventral cord or neurite data) - `0`
- out path - `Results` (path should have forward slashes '/')
- architecture - `unet_fixed`
- training mode - `2D`
- loss - `l1`
- number of training epochs - `200`
- batch size - `10`
run following commands -
1) `env\Scripts\activate.bat`
2) `python train.py /training_data 1 0 -out Results -arch unet_fixed -mode 2D -loss l1 -epoch 200 -bs 10`
4. Once training is finished a folder named `run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]` will be created in `Results` folder. Output files in this folder should look like below (e.g. shown for a sample run)
```
Results
└───run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]
checkpoint
events.out.tfevents.1618996150.atl1-1-01-004-33.pace.gatech.edu
model.data-00000-of-00001
model.index
model.meta
pred_1999.png
pred_2180.png
pred_2227.png
pred_2492.png
pred_335.png
test_data_loss.txt
training_loss.txt
X_1999.png
X_2180.png
X_2227.png
X_2492.png
X_335.png
Y_1999.png
Y_2180.png
Y_2227.png
Y_2492.png
Y_335.png
```
Here -
- `training_loss.txt` stores epoch wise loss information on randomly sampled images from training data and test data.
- `test_data_loss.txt` stores loss on randomly sampled images from test data.
- `checkpoint`, `events*` and `model*` files will be used to restore trained weights of network to perform inference on new data.
- `X*.png`, `Y*.png` denote randomly selected noisy (input) and clean (ground-truth) images from test data. `pred*.png` denote corresponding denoised predictions by trained network.
# Denoise calcium activity recordings
1. Structure of functional recording video datasets should be organised as below -
```
vid1
└───noisy_imgs
├───img_1
│ z_1.tif
│ z_2.tif
│ ...
│
├───img_2
│ z_1.tif
│ z_2.tif
│ ...
│
├───img_3
│ z_1.tif
│ z_2.tif
│ ...
...
```
Here `img_1`, `img_2` etc. can correspond to individual time-points in videos.
2. Run `python inference.py -h` to see usage and input arguments. Output on terminal should look like below -
```
usage: inference.py [-h] data [data ...] run
denoise images using weights of trained model
positional arguments:
data paths of datasets to be denoised
run path of saved trained model
optional arguments:
-h, --help show this help message and exit
```
e.g to denoise two video datasets with following settings -
- two video datasets with paths - `/vid1`, `/vid2` (path should have forward slashes '/')
- save trained network path - `Results/run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]` (path should have forward slashes '/')
run following commands -
1) `env\Scripts\activate.bat`
2) `python inference.py /vid1 /vid2 /Results/run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]`
__Note :__ to denoise max-projection calcium imaging datasets, folder structure of input dataset should remain same. In this case, use a network trained on max-projection images to denoise e.g. `Results/run_hourglass_wres_l1_mp1_m2D_d1_1_[tsize]` (here mp1 in run name denotes network was trained on max-projection images).
3. Once denoising is finished, output folders `/vid1/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]` and `/vid2/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]` will be created. Output files in these folders should look like below -
```
/vid1/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]
│ run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_inference_runtime.txt
│
├───img_1
│ z_1.tif
│ z_2.tif
│ ...
│
├───img_2
│ z_1.tif
│ z_2.tif
│ ...
│
├───img_3
│ z_1.tif
│ z_2.tif
│ ...
...
```
Here -
- `run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_inference_runtime.txt` stores inference runtime information for each image.
- `img_1`, `img_2` etc. folders store denoised images corresponding to noisy video.
# Denoise independent images
1. E.g. to denoise 3 individual images (may be 2D images or 3D stacks) with following settings -
- 3 images with paths - `data1/img_1.tif`, `data2/img_2.png`, `data2/img_3.png` (path should have forward slashes '/')
- save trained network path - `Results/run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]` (path should have forward slashes '/')
run following commands -
1) `env\Scripts\activate.bat`
2) `python inference.py data1/img_1.tif data2/img_2.png data2/img_3.png Results/run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]`
2. Once denoising is finished, output folders `data1/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_img_1.tif`, `data2/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_img_2.png` and `data2/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_img_3.png` will be created. Output files in these folders should look like below -
```
data2/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_img_2.png
run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_inference_runtime.txt
img_2.png.tif
```
Here -
- `run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_inference_runtime.txt` stores inference runtime information for image.
- `img_2.png.tif` is the denoised output corresponding to `data2/img_2.png`
# Run sample datasets
For demo, we provide three sample datasets, models trained for these datasets, and denoising outputs for these datasets in `sample_runs` folder. These datasets include -
- `Whole_brain` - z-planes from low SNR whole-brain image stacks
- `Mechanosensory` - max-projection images from low SNR stacks of neurites of mechanosensory neurons
- `VentralCord` - max-projection images from low SNR stacks of motor neurons in ventral cord.
Structure of these folders looks like below (shown for `\sample_runs\VentralCord`) -
```
\sample_runs\VentralCord
│ X_101.png
│ X_109.png
│ X_11.png
│ X_115.png
│ X_118.png
│ X_122.png
│ X_123.png
│ X_26.png
│ X_42.png
│ X_67.png
│ X_76.png
│ X_81.png
│ X_91.png
│ X_92.png
│
├───run_hourglass_wres_l2_mp1_m2D_d1_2_317
│ checkpoint
│ events.out.tfevents.1625789062.atl1-1-02-006-35.pace.gatech.edu
│ model.data-00000-of-00001
│ model.index
│ model.meta
│ test_data_loss.txt
│ training_loss.txt
│
└───sample_output
├───pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_101.png
│ run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt
│ X_101.png.tif
│
├───pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_11.png
│ run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt
│ X_11.png.tif
│
├───pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_122.png
│ run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt
│ X_122.png.tif
│
├───pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_42.png
│ run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt
│ X_42.png.tif
│
├───pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_81.png
│ run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt
│ X_81.png.tif
│
├───pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_91.png
│ run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt
│ X_91.png.tif
│
└───pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_92.png
run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt
X_92.png.tif
```
Here -
- `X_*.png` correspond to low SNR images.
- `run_hourglass_wres_l2_mp1_m2D_d1_2_317` stores `hourglass_wres` cnn trained on max-projection images (`mp1`) from ventral cord dataset in `2D` mode using `l2` loss function. The stored model can be used to denoise images.
- `sample_output\pred_*` folders store denoised outputs of trained cnn for corresponding `X_*.png` images.
To denoise an example image, run following commands -
1) `env\Scripts\activate.bat`
2) `python inference.py sample_runs/VentralCord/X_26.png sample_runs/VentralCord/run_hourglass_wres_l2_mp1_m2D_d1_2_317`
Upon completion, a new folder, `pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_26.png`, will be created in the same directory that contains denoised output. The folder structure should like below now -
```
\sample_runs\VentralCord
│ X_101.png
│ X_109.png
│ X_11.png
│ X_115.png
│ X_118.png
│ X_122.png
│ X_123.png
│ X_26.png
│ X_42.png
│ X_67.png
│ X_76.png
│ X_81.png
│ X_91.png
│ X_92.png
│
├───pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_26.png
│ run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt
│ X_26.png.tif
│
├───run_hourglass_wres_l2_mp1_m2D_d1_2_317
│ checkpoint
│ events.out.tfevents.1625789062.atl1-1-02-006-35.pace.gatech.edu
│ model.data-00000-of-00001
│ model.index
│ model.meta
│ test_data_loss.txt
│ training_loss.txt
│
└───sample_output
```
Raw data
{
"_id": null,
"home_page": "https://github.com/shiveshc/whole-brain_DeepDenoising",
"name": "NIDDL",
"maintainer": "",
"docs_url": null,
"requires_python": "",
"maintainer_email": "",
"keywords": "denoising,whole-brain imaging,calcium imaging,celegans,deep learning,volumetric imaging,microscopy",
"author": "Shivesh Chaudhary",
"author_email": "shiveshc@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/e9/42/3e200516459eb3c53d97f940214f86f6b511f8e74e88848367e761c34467/NIDDL-0.1.8.tar.gz",
"platform": null,
"description": "\n\n\n\n\n\n<!--  -->\n\n\n# whole-brain_DeepDenoising\nDeep denoising pushes the limit of functional data acquisition by recovering high SNR calcium traces from low SNR videos acquired using low laser power or smaller exposure time. Thus deep denoising enables faster and longer volumetric recordings. For more details, our [paper](https://www.nature.com/articles/s41467-022-32886-w) is now published.\n\n<p align = \"center\"><b>Denoise whole-brain videos</b></p>\n\n<!-- <img src = \"https://github.com/shiveshc/whole-brain_DeepDenoising/blob/main/extra/wb_denoising.gif\" width=100% align=\"center\"> -->\n\n<p align = \"center\"><b>Denoise mechanosensory neurites</b></p>\n\n<!-- <img src = \"https://github.com/shiveshc/whole-brain_DeepDenoising/blob/main/extra/neurite_denoising.gif\" width=100% align=\"center\"> -->\n\n<p align = \"center\"><b>Recover high SNR calcium traces</b></p>\n\n<!-- <img src = \"https://github.com/shiveshc/whole-brain_DeepDenoising/blob/main/extra/denoise_traces.png\" width=100% align=\"center\"> -->\n\n# Contents\n1. [Installation](#installation)\n2. [Additional system requirements](#additional-system-requirements)\n3. [Train on new dataset](#train-on-new-dataset)\n4. [Denoise calcium activity recordings](#denoise-calcium-activity-recordings)\n5. [Denoise independent images](#denoise-independent-images)\n6. [Run sample datasets](#run-sample-datasets)\n\n# Installation\nInstallation steps tested for Windows 10 64-bit and Python 3.5\n\n### 1. Download whole-brain_DeepDenoising repository\nOpen Git Bash terminal, navigate to desired location and clone repository using `git clone https://github.com/shiveshc/whole-brain_DeepDenoising.git`.\n\nOr click on `Code` button on top right corner and `Download ZIP`.\n\n### 2. Setting up venv and installing libraries\nOpen command line terminal as administrator and navigate to cloned repository path using `cd .\\whole-brain_DeepDenoising`.\n\nNext run following commands - \n| Windows 10 | MacOS |\n|------------|-------|\n| 1. `python -m venv env` | 1. `python -m venv env` |\n| 2. `env\\Scripts\\activate.bat` | 2. `source env/bin/activate` |\n| 3. `python -m pip install --upgrade \"pip < 21.0\"` | 3. `python -m pip install --upgrade \"pip < 21.0\"` |\n| 4. `pip install -r requirements.txt` | 4. `pip install -r requirements.txt` |\n\nInstallation should take 10-15 minutes.\n\nCommon installation errors -\n1. pip version is not upgraded\\\n __solution :__ upgrade pip using `python -m pip install --upgrade pip`\n2. pip version is not compatible with python version\\\n __solution :__ install suitable pip version e.g. pip version < 21.0 is compatible with Python 3.5.\n \n# Additional system requirements\nModel training using GPUs is much faster and thus preferred. To be able to use GPUs, suitable NVIDIA drivers and CUDA libraries must be installed. Additional instructions on setting up tensorflow-gpu can be found at <a href = \"https://www.tensorflow.org/install/gpu#Software_requirements\">Software requirements</a> and <a href = \"https://www.tensorflow.org/install/gpu#windows_setup\">Windows setup</a>.\n\nFor tensorflow 1.6.0 currently setup in venv, CUDA v9.0 and `cudnn-9.0-windows10-x64-v7.6.4.38` downloaded from <a href= \"https://developer.nvidia.com/rdp/cudnn-archive\">cudnn archives</a> works for Windows 10 64-bit.\n\n# Train on new dataset\n1. To train network on new data, pairs of noisy i.e. low SNR images (acquired at low laser power or small exposure time conditions) and clean i.e. high SNR images (acquired at high laser power or long exposure time conditions) are needed. Currently supported image size is 512 x 512 x d where d is number of images in stack. For other image sizes, either resize images first or channel dimensions need to be changed in architecture files in `cnn_archs` folder.\n\n2. Structure of training data folder should be organised as below -\n ```\n data\n \u251c\u2500\u2500\u2500gt_imgs\n \u2502 \u251c\u2500\u2500\u2500img_1\n \u2502 \u2502 z_1.tif\n \u2502 \u2502 z_2.tif\n \u2502 \u2502 ...\n \u2502 \u2502\n \u2502 \u251c\u2500\u2500\u2500img_2\n \u2502 \u2502 z_1.tif\n \u2502 \u2502 z_2.tif\n \u2502 \u2502 ...\n \u2502 \u2502\n \u2502 \u2514\u2500\u2500\u2500img_3\n \u2502 z_1.tif\n \u2502 z_2.tif\n \u2502 ...\n \u2502\n \u2514\u2500\u2500\u2500noisy_imgs\n \u251c\u2500\u2500\u2500img_1\n \u2502 z_1.tif\n \u2502 z_2.tif\n \u2502 ...\n \u2502\n \u251c\u2500\u2500\u2500img_2\n \u2502 z_1.tif\n \u2502 z_2.tif\n \u2502 ...\n \u2502\n \u2514\u2500\u2500\u2500img_3\n z_1.tif\n z_2.tif\n ...\n ```\n \n 3. Run `python train.py -h` to see usage and input arguments. The output on terminal should look like below -\n ```\n usage: train.py [-h] [-out OUT]\n [-arch {unet,unet_fixed,hourglass_wres,hourglass_wores}]\n [-mode {2D,2.5D,3D}] [-depth DEPTH] [-loss {l2,l1}]\n [-epochs EPOCHS] [-lr LR] [-bs BS] [-tsize TSIZE]\n data run {1,0}\n\n train CNN model to denoise volumetric functional recordings\n\n positional arguments:\n data training data path\n run run number to distinguish different runs\n {1,0} 1 if train network on max projection of 3D stacks else\n 0\n\n optional arguments:\n -h, --help show this help message and exit\n -out OUT location for saving results\n -arch {unet,unet_fixed,hourglass_wres,hourglass_wores}\n CNN architechture to use for training (default is\n hourglass_wres)\n -mode {2D,2.5D,3D} training mode (default is 2D)\n -depth DEPTH stack depth to use for training (must be odd number,\n default is 1)\n -loss {l2,l1} L2 or L1 loss for training (default is l1)\n -epochs EPOCHS number of epochs to train the model for (150-200 is\n good choice, default is 150)\n -lr LR learning rate (default is 0.001)\n -bs BS batch size of training (default is 6)\n -tsize TSIZE data size (number of images) to use for training\n ```\n e.g. to train the network on whole-brain data with following settings -\n - data path - `/training_data` (path should have forward slashes '/')\n - run number - `1`\n - train on max-projection (e.g. ventral cord or neurite data) - `0`\n - out path - `Results` (path should have forward slashes '/')\n - architecture - `unet_fixed`\n - training mode - `2D`\n - loss - `l1`\n - number of training epochs - `200`\n - batch size - `10`\n\n run following commands -\n 1) `env\\Scripts\\activate.bat`\n 2) `python train.py /training_data 1 0 -out Results -arch unet_fixed -mode 2D -loss l1 -epoch 200 -bs 10`\n \n 4. Once training is finished a folder named `run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]` will be created in `Results` folder. Output files in this folder should look like below (e.g. shown for a sample run)\n ```\n Results\n \u2514\u2500\u2500\u2500run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]\n checkpoint\n events.out.tfevents.1618996150.atl1-1-01-004-33.pace.gatech.edu\n model.data-00000-of-00001\n model.index\n model.meta\n pred_1999.png\n pred_2180.png\n pred_2227.png\n pred_2492.png\n pred_335.png\n test_data_loss.txt\n training_loss.txt\n X_1999.png\n X_2180.png\n X_2227.png\n X_2492.png\n X_335.png\n Y_1999.png\n Y_2180.png\n Y_2227.png\n Y_2492.png\n Y_335.png\n ```\n Here -\n - `training_loss.txt` stores epoch wise loss information on randomly sampled images from training data and test data.\n - `test_data_loss.txt` stores loss on randomly sampled images from test data.\n - `checkpoint`, `events*` and `model*` files will be used to restore trained weights of network to perform inference on new data.\n - `X*.png`, `Y*.png` denote randomly selected noisy (input) and clean (ground-truth) images from test data. `pred*.png` denote corresponding denoised predictions by trained network.\n\n\n# Denoise calcium activity recordings\n1. Structure of functional recording video datasets should be organised as below - \n ```\n vid1\n \u2514\u2500\u2500\u2500noisy_imgs\n \u251c\u2500\u2500\u2500img_1\n \u2502 z_1.tif\n \u2502 z_2.tif\n \u2502 ...\n \u2502\n \u251c\u2500\u2500\u2500img_2\n \u2502 z_1.tif\n \u2502 z_2.tif\n \u2502 ...\n \u2502\n \u251c\u2500\u2500\u2500img_3\n \u2502 z_1.tif\n \u2502 z_2.tif\n \u2502 ...\n ...\n ```\n Here `img_1`, `img_2` etc. can correspond to individual time-points in videos.\n\n2. Run `python inference.py -h` to see usage and input arguments. Output on terminal should look like below - \n ```\n usage: inference.py [-h] data [data ...] run\n\n denoise images using weights of trained model\n\n positional arguments:\n data paths of datasets to be denoised\n run path of saved trained model\n\n optional arguments:\n -h, --help show this help message and exit\n ```\n\n e.g to denoise two video datasets with following settings - \n - two video datasets with paths - `/vid1`, `/vid2` (path should have forward slashes '/')\n - save trained network path - `Results/run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]` (path should have forward slashes '/')\n\n run following commands -\n 1) `env\\Scripts\\activate.bat`\n 2) `python inference.py /vid1 /vid2 /Results/run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]`\n\n __Note :__ to denoise max-projection calcium imaging datasets, folder structure of input dataset should remain same. In this case, use a network trained on max-projection images to denoise e.g. `Results/run_hourglass_wres_l1_mp1_m2D_d1_1_[tsize]` (here mp1 in run name denotes network was trained on max-projection images).\n\n3. Once denoising is finished, output folders `/vid1/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]` and `/vid2/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]` will be created. Output files in these folders should look like below -\n ```\n /vid1/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]\n \u2502 run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_inference_runtime.txt\n \u2502\n \u251c\u2500\u2500\u2500img_1\n \u2502 z_1.tif\n \u2502 z_2.tif\n \u2502 ...\n \u2502\n \u251c\u2500\u2500\u2500img_2\n \u2502 z_1.tif\n \u2502 z_2.tif\n \u2502 ...\n \u2502\n \u251c\u2500\u2500\u2500img_3\n \u2502 z_1.tif\n \u2502 z_2.tif\n \u2502 ...\n ...\n ```\n Here -\n - `run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_inference_runtime.txt` stores inference runtime information for each image.\n - `img_1`, `img_2` etc. folders store denoised images corresponding to noisy video.\n\n# Denoise independent images\n1. E.g. to denoise 3 individual images (may be 2D images or 3D stacks) with following settings - \n - 3 images with paths - `data1/img_1.tif`, `data2/img_2.png`, `data2/img_3.png` (path should have forward slashes '/')\n - save trained network path - `Results/run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]` (path should have forward slashes '/')\n\n run following commands -\n 1) `env\\Scripts\\activate.bat`\n 2) `python inference.py data1/img_1.tif data2/img_2.png data2/img_3.png Results/run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]`\n\n2. Once denoising is finished, output folders `data1/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_img_1.tif`, `data2/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_img_2.png` and `data2/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_img_3.png` will be created. Output files in these folders should look like below -\n ```\n data2/pred_run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_img_2.png\n run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_inference_runtime.txt\n img_2.png.tif\n ```\n Here -\n - `run_unet_fixed_l1_mp0_m2D_d1_1_[tsize]_inference_runtime.txt` stores inference runtime information for image.\n - `img_2.png.tif` is the denoised output corresponding to `data2/img_2.png`\n\n# Run sample datasets\nFor demo, we provide three sample datasets, models trained for these datasets, and denoising outputs for these datasets in `sample_runs` folder. These datasets include -\n- `Whole_brain` - z-planes from low SNR whole-brain image stacks\n- `Mechanosensory` - max-projection images from low SNR stacks of neurites of mechanosensory neurons\n- `VentralCord` - max-projection images from low SNR stacks of motor neurons in ventral cord.\n\nStructure of these folders looks like below (shown for `\\sample_runs\\VentralCord`) - \n```\n\\sample_runs\\VentralCord\n\u2502 X_101.png\n\u2502 X_109.png\n\u2502 X_11.png\n\u2502 X_115.png\n\u2502 X_118.png\n\u2502 X_122.png\n\u2502 X_123.png\n\u2502 X_26.png\n\u2502 X_42.png\n\u2502 X_67.png\n\u2502 X_76.png\n\u2502 X_81.png\n\u2502 X_91.png\n\u2502 X_92.png\n\u2502\n\u251c\u2500\u2500\u2500run_hourglass_wres_l2_mp1_m2D_d1_2_317\n\u2502 checkpoint\n\u2502 events.out.tfevents.1625789062.atl1-1-02-006-35.pace.gatech.edu\n\u2502 model.data-00000-of-00001\n\u2502 model.index\n\u2502 model.meta\n\u2502 test_data_loss.txt\n\u2502 training_loss.txt\n\u2502\n\u2514\u2500\u2500\u2500sample_output\n \u251c\u2500\u2500\u2500pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_101.png\n \u2502 run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt\n \u2502 X_101.png.tif\n \u2502\n \u251c\u2500\u2500\u2500pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_11.png\n \u2502 run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt\n \u2502 X_11.png.tif\n \u2502\n \u251c\u2500\u2500\u2500pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_122.png\n \u2502 run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt\n \u2502 X_122.png.tif\n \u2502\n \u251c\u2500\u2500\u2500pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_42.png\n \u2502 run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt\n \u2502 X_42.png.tif\n \u2502\n \u251c\u2500\u2500\u2500pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_81.png\n \u2502 run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt\n \u2502 X_81.png.tif\n \u2502\n \u251c\u2500\u2500\u2500pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_91.png\n \u2502 run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt\n \u2502 X_91.png.tif\n \u2502\n \u2514\u2500\u2500\u2500pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_92.png\n run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt\n X_92.png.tif\n```\nHere -\n- `X_*.png` correspond to low SNR images.\n- `run_hourglass_wres_l2_mp1_m2D_d1_2_317` stores `hourglass_wres` cnn trained on max-projection images (`mp1`) from ventral cord dataset in `2D` mode using `l2` loss function. The stored model can be used to denoise images.\n- `sample_output\\pred_*` folders store denoised outputs of trained cnn for corresponding `X_*.png` images.\n\nTo denoise an example image, run following commands -\n1) `env\\Scripts\\activate.bat`\n2) `python inference.py sample_runs/VentralCord/X_26.png sample_runs/VentralCord/run_hourglass_wres_l2_mp1_m2D_d1_2_317`\n\nUpon completion, a new folder, `pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_26.png`, will be created in the same directory that contains denoised output. The folder structure should like below now -\n```\n\\sample_runs\\VentralCord\n\u2502 X_101.png\n\u2502 X_109.png\n\u2502 X_11.png\n\u2502 X_115.png\n\u2502 X_118.png\n\u2502 X_122.png\n\u2502 X_123.png\n\u2502 X_26.png\n\u2502 X_42.png\n\u2502 X_67.png\n\u2502 X_76.png\n\u2502 X_81.png\n\u2502 X_91.png\n\u2502 X_92.png\n\u2502\n\u251c\u2500\u2500\u2500pred_run_hourglass_wres_l2_mp1_m2D_d1_2_317_X_26.png\n\u2502 run_hourglass_wres_l2_mp1_m2D_d1_2_317_inference_runtime.txt\n\u2502 X_26.png.tif\n\u2502\n\u251c\u2500\u2500\u2500run_hourglass_wres_l2_mp1_m2D_d1_2_317\n\u2502 checkpoint\n\u2502 events.out.tfevents.1625789062.atl1-1-02-006-35.pace.gatech.edu\n\u2502 model.data-00000-of-00001\n\u2502 model.index\n\u2502 model.meta\n\u2502 test_data_loss.txt\n\u2502 training_loss.txt\n\u2502\n\u2514\u2500\u2500\u2500sample_output\n```\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "Neuro Imaging Denoising via Deep Learning (NIDDL)",
"version": "0.1.8",
"project_urls": {
"Homepage": "https://github.com/shiveshc/whole-brain_DeepDenoising"
},
"split_keywords": [
"denoising",
"whole-brain imaging",
"calcium imaging",
"celegans",
"deep learning",
"volumetric imaging",
"microscopy"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "3718f581c9683aed779cce91e086c13c538d5287c4b34b6390558fcc148af637",
"md5": "94f87923e687425035ccc77312575358",
"sha256": "089a8899319fd98375da40af60acc05a9efa6e0deb945feb50bfa802f79c9650"
},
"downloads": -1,
"filename": "NIDDL-0.1.8-py3-none-any.whl",
"has_sig": false,
"md5_digest": "94f87923e687425035ccc77312575358",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": null,
"size": 56232,
"upload_time": "2023-07-03T18:48:52",
"upload_time_iso_8601": "2023-07-03T18:48:52.164704Z",
"url": "https://files.pythonhosted.org/packages/37/18/f581c9683aed779cce91e086c13c538d5287c4b34b6390558fcc148af637/NIDDL-0.1.8-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "e9423e200516459eb3c53d97f940214f86f6b511f8e74e88848367e761c34467",
"md5": "a3f3b36badaa86250cb7e02b4878d209",
"sha256": "297a9478c4cfffa41de235bc22a82d86834dabd7823546b864680796262fffe6"
},
"downloads": -1,
"filename": "NIDDL-0.1.8.tar.gz",
"has_sig": false,
"md5_digest": "a3f3b36badaa86250cb7e02b4878d209",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 29325,
"upload_time": "2023-07-03T18:48:54",
"upload_time_iso_8601": "2023-07-03T18:48:54.002317Z",
"url": "https://files.pythonhosted.org/packages/e9/42/3e200516459eb3c53d97f940214f86f6b511f8e74e88848367e761c34467/NIDDL-0.1.8.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-07-03 18:48:54",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "shiveshc",
"github_project": "whole-brain_DeepDenoising",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [],
"lcname": "niddl"
}
JSON
